Validation and performance of the sleep inertia questionnaire in central disorders of hypersomnolence

Optimal measurement tools for problematic sleep inertia, common in some central disorders of hypersomnolence (CDH), have not yet been determined. We evaluated the performance of the Sleep Inertia Questionnaire (SIQ) in CDH, and how well it distinguished hypersomnolent groups from controls, and IH (idiopathic hypersomnia) from narcolepsy type 1 (NT1).

This prospective, bi-centric study included 63 control, 84 IH, 16 NT1, 18 narcolepsy type 2 (NT2), and 88 subjective excessive daytime sleepiness (sEDS) participants, using ICSD-3 criteria. 126 (47.2 %) participants were on any medication at the time of SIQ completion. We assessed construct validity of SIQ scores, and sleep inertia duration (SID), and compared them across diagnoses, controlling for age and center. We derived cutpoints to distinguish hypersomnolent patients from controls and IH from NT1. Sensitivity analyses for depression, chronotype, and medication were performed.

The SIQ sum and composite score were significantly lower in controls than in other groups (p < 0.0001), demonstrating outstanding ability to distinguish patients from controls (AUCs 0.92), without differences among hypersomnolent groups. SID (AUC 0.76) was significantly shorter in controls than in all hypersomnolent groups except NT1, and was shorter in NT1 than in IH or sEDS. Optimal SIQ sum cutpoint was 42 (J = 0.71) for patients versus controls. Optimal SID cutpoint in distinguishing IH from NT1 was 25 min (J = 0.39).

The SIQ has excellent ability to distinguish hypersomnolent patients from healthy controls, after controlling for depression, eveningness, and medication. SID is best at distinguishing IH from NT1.